Skip to main content
. 2021 Jan 19;23:9. doi: 10.1186/s13058-021-01386-y

Table 2.

Clinical and laboratory features at MAHA diagnosis and association with 4-week survival

Univariate analysis Multivariate analysis
Factors N (%) OR 95% IC p value OR 95% IC p value
PS at MAHA diagnosis
 PS 1–2 30 (55.6%) 1.0 0.003 1.0 0.01
 PS 3–4 24 (44.4%) 6.0 [1.8; 19.8] 7.0 [1.6; 31.8]
Clinical bleeding
 Yes 9 (16.7%) NS NS
 No 45 (83.3%)
Neurological symptoms
 Yes 15 (28.3%) NS NS
 No 38 (71.7%)
 NA 1
Dyspnoea
 Yes 11 (20.8%) NS NS
 No 42 (79.2%)
 NA 1
Metastatic sites
 < 3 25 (46.3%) NS NS
 ≥ 3 29 (53.7%)
Metastasis
 Bone marrowa 11 (78.5%)
 Bone 44 (81.5%)
 Lung 8 (14.8%) NS NS
 Liver 35 (64.8%)
 Otherb 24 (44.4%)
Number of prior treatment lines
 ≤ 1 26 (48.1%) 1.0 0.06 NS
 > 1 28 (51.9%) 2.9 [1.0; 8.7]
Platelets
 < 50 G/L 29 (53.7%) NS NS
 ≥ 50 G/L 25 (46.3%)
Haemoglobin
 < 8 g/L 26 (48.1%) 4.0 [1.3; 12.5] 0.02 3.7 [0.9; 16.7] 0.08
 ≥ 8 g/L 28 (51.9%) 1.0 1.0
Schistocytes
 > 5.0% 10 (25.0%) NS NS
 0.5–5.0% 30 (75.0%)
 NA 14
Erythroblastemia
 Yes 40 (85.1%) NS NS
 No 7 (14.9%)
 NA 7
Myelemia
 Yes 38 (90.5%) NS NS
 No 4 (9.5%)
 NA 12
Prothrombin time
 < 50% 8 (16.0%) 4.5 [0.9; 25.0] 0.07 9.1 [1.2; 50.0] 0.03
 ≥ 50% 42 (84.0%) 1.0 1.0
 NA 4
Fibrinogen
 ≤ 2 g/L 15 (34.9) NS NS
 > 2 g/L 28 (65.1)
 NA 11
Glomerular filtration eate
 > 60 mL/min 25 (65.8%) NS NS
 30–60 mL/min 9 (23.7%) NS
 < 30 mL/min 4 (10.5%) NS
 NA 16
Total bilirubin level
 < 1.24 mg/dL 16 (30.8%) 1.0 0.007 1.0
 ≥ 1.24 mg/dL 36 (69.2%) 7.3 [1.8; 30.6] 6.9 [1.1; 42.6] 0.04
 NA 2
Pronopall scorec
 Short survival 12 (31.6%) 3.7 [0.8; 16.8] 0.09
 Intermediate/long survival 26 (68.4%) 1.0 NS
 NA 16

NA not available, OR odds ratio, NS non-significant

aOnly 14 patients had a bone marrow examination (myelogram or bone marrow biopsy) at the time of BC-MAHA diagnosis

bOthers metastatic sites: cerebral, carcinomatous meningitis, node involvement

cThe pronopall score for early death among oncology patients was calculated according to Barbot et al., J Clin Oncol 2008, missing data being imputed